[Dataset] Fix CoLA dataset index and add Glue dataset API

[eric-haibin-lin]

Description

Added Glue datasets to the dataset API. Will refactor scripts/bert/dataset.py in a followup PR.

Checklist

Essentials

• PR's title starts with a category (e.g. [BUGFIX], [MODEL], [TUTORIAL], [FEATURE], [DOC], etc)
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage
• Code is well-documented

Changes

• Feature1, tests, (and when applicable, API doc)
• Feature2, tests, (and when applicable, API doc)

Comments

• If this change is a backward incompatible change, why must this change be made.
• Interesting edge cases to note here

[codecov]

Codecov Report

Merging #682 into master will increase coverage by 0.9%.
The diff coverage is 98.91%.

@@            Coverage Diff            @@
##           master     #682     +/-   ##
=========================================
+ Coverage   64.48%   65.38%   +0.9%     
=========================================
  Files         154      143     -11     
  Lines       14385    13269   -1116     
=========================================
- Hits         9276     8676    -600     
+ Misses       5109     4593    -516

Flag	Coverage Δ
#PR435	?
#PR466	?
#PR588	?
#PR612	?
#PR648	?
#PR663	?
#PR669	?
#PR672	?
#PR678	?
#PR682	65.38% <98.91%> (?)
#master	?
#notserial	41.71% <28.64%> (+2.89%)	⬆️
#py2	65.21% <98.91%> (-0.04%)	⬇️
#py3	65.63% <28.64%> (+1.27%)	⬆️
#serial	51.4% <98.91%> (+0.21%)	⬆️

[codecov]

Codecov Report

Merging #682 into master will increase coverage by 0.47%.
The diff coverage is 98.91%.

@@            Coverage Diff             @@
##           master     #682      +/-   ##
==========================================
+ Coverage   64.98%   65.45%   +0.47%     
==========================================
  Files         143      144       +1     
  Lines       13127    13300     +173     
==========================================
+ Hits         8530     8706     +176     
+ Misses       4597     4594       -3

Flag	Coverage Δ
#PR682	65.45% <98.91%> (?)
#master	?
#notserial	41.71% <28.64%> (+0.13%)	⬆️
#py2	65.21% <98.91%> (-23.89%)	⬇️
#py3	65.34% <98.91%> (+0.44%)	⬆️
#serial	51.61% <98.91%> (+0.44%)	⬆️

[mli]

Job PR-682/13 is complete.
Docs are uploaded to http://gluon-nlp-staging.s3-accelerate.dualstack.amazonaws.com/PR-682/13/index.html

[eric-haibin-lin]

@Gpwner could you help review the CoLADataset class?

[vanewu]

Nice work. Every data is correct. When checking this, I found that the field of the test data in the previous COLADataset was incorrect. https://github.com/dmlc/gluon-nlp/blob/master/scripts/bert/dataset.py#L300.